U.S. patent number 5,032,978 [Application Number 07/414,211] was granted by the patent office on 1991-07-16 for status tree monitoring and display system.
This patent grant is currently assigned to Westinghouse Electric Co.. Invention is credited to John P. Carrera, James R. Easter, Mary C. Eastman, William C. Elm, Melvin H. Lipner, A. Dean Mundy, Craig D. Watson, David D. Woods.
United States Patent |
5,032,978 |
Watson , et al. |
July 16, 1991 |
**Please see images for:
( Certificate of Correction ) ** |
Status tree monitoring and display system
Abstract
A display, a display method and an apparatus are disclosed which
produce a summary display depicting function states using discrete
state bars centered in one window of a two-window display. The
other window on the summary display includes meters for the
parameters along the active path of the status tree for the most
relevant function as determined by the operator. Each meter
indicates not only the current value of the parameter being
monitored but the ranges of the different function states as they
correspond to the parameter being displayed by the meter. Each
meter also carries a data quality indicator. A second level display
is provided which depicts the status tree with the active path
highlighted allowing the operator to review the decision made by
the status tree. The second level display also indicates the values
of the parameters used in the decision-making process and includes
a miniature version of the summary display that allows the operator
to continuously monitor system summary status while reviewing the
second level display. A third level display is also provided which
depicts detailed information on the sensor values of the parameters
being monitored as well as the exact questions being answered by
decision trees of the process control monitoring system along with
the answers. This third level display also includes a miniature
version of the summary status display.
Inventors: |
Watson; Craig D. (Monroeville,
PA), Eastman; Mary C. (Pittsburgh, PA), Woods; David
D. (Murrysville, PA), Carrera; John P. (Greensburg,
PA), Easter; James R. (Pittsburgh, PA), Lipner; Melvin
H. (Monroeville, PA), Elm; William C. (Monroeville,
PA), Mundy; A. Dean (Gibsonia, PA) |
Assignee: |
Westinghouse Electric Co.
(Pittsburgh, PA)
|
Family
ID: |
27396372 |
Appl.
No.: |
07/414,211 |
Filed: |
September 28, 1989 |
Related U.S. Patent Documents
|
|
|
|
|
|
|
Application
Number |
Filing Date |
Patent Number |
Issue Date |
|
|
217117 |
Jul 7, 1988 |
4902469 |
|
|
|
859406 |
May 5, 1986 |
|
|
|
|
Current U.S.
Class: |
700/83; 376/217;
976/DIG.207 |
Current CPC
Class: |
G05B
23/0227 (20130101); G21C 17/00 (20130101); G05B
23/0272 (20130101); Y02E 30/30 (20130101) |
Current International
Class: |
G05B
23/02 (20060101); G21C 17/00 (20060101); G06F
015/46 (); G21C 007/36 () |
Field of
Search: |
;364/188,189,146,138,139,152-157 ;376/215,216,217,245,248,259
;340/722,753 |
References Cited
[Referenced By]
U.S. Patent Documents
Foreign Patent Documents
|
|
|
|
|
|
|
0083546 |
|
Jul 1983 |
|
EP |
|
0099681 |
|
Feb 1984 |
|
EP |
|
59-109914 |
|
Dec 1982 |
|
JP |
|
60-263215 |
|
Dec 1985 |
|
JP |
|
Primary Examiner: Ruggiero; Joseph
Parent Case Text
This is a division of application Ser. No. 07/217,117, filed July
7, 1988, now U.S. Pat. No. 4,902,469, which is a continuation of
Ser. No. 859,406 filed May 5, 1986, abandoned.
Claims
What is claimed is:
1. A display for a process control system monitoring process
control data which determines discrete process parameters, said
display comprising:
a meter image including an actual value indicator indicating an
actual parameter value of a portion of the process control data and
a scale spatially associated with the actual value indicator;
and
variable range brackets spatially associated with said meter image
each indicating ranges for the actual parameter value, said
brackets are color coded in dependence on a system status which
results when the parameter value falls within the bracket
range.
2. A display for a process control system monitoring process
control data which determines discrete process parameters, said
display comprising:
a meter image including an actual value indicator indicating an
actual parameter value of a portion of the process control data and
a scale spatially associated with the actual value indicator;
variable range brackets spatially associated with said meter image
each indicating ranges for the actual parameter value; and
a procedure name spatially associated with each bracket, the
procedure name indicating a procedure performed by an operator in
dependence on the actual parameter value.
3. A display for a process control system monitoring process
control data which determines discrete process parameters, said
display comprising:
meter images, each meter image associated with at least one item
along an active path in a status tree designating the parameter,
each meter image identifies the item monitored and provides a
digital value for the item and each meter image including an actual
value indicator indicating an actual value of a portion of the
process control data and a scale spatially associated with the
actual value indicator; and
variable range brackets spatially associated with each said meter
image each indicating process status ranges for the actual
value.
4. A display for a process control system monitoring process
control data which determines discrete process parameters, said
display comprising:
a meter image including an actual value status indicator indicating
an actual value of a portion of the process control data, the
actual value status indicator including a data quality indicator
and a scale spatially associated with the actual value indicator;
and
variable range brackets spatially associated with said meter image
each indicating ranges for the actual value.
5. A display for a process control system monitoring process
control data which determines discrete process parameters, said
display comprising:
a meter image including an actual value indicator indicating an
actual value of a portion of the process control data and a scale
spatially associated with the actual value indicator; and
variable range brackets spatially associated with said meter image
each indicating ranges for the actual value where the process has
discrete states and the each range indicted by each bracket
indicates values of the parameter for which the state of the
process will remain the same.
6. A display, comprising:
a meter image including an actual value indicator indicating an
actual value of data and a scale spatially associated with the
actual value indicator; and
variable range brackets spatially associated with said meter image
each indicating a process status range for the actual value, the
actual value of the data indicates the actual value of a process
which has discrete states and the range indicated by each bracket
indicates values of the parameter for which the status of the
process will remain the same.
7. A display as recited in claim 3, wherein each range bracket
associated with one of the meter images indicates a process status
determined from parameters in addition to a parameter associated
with the actual value indicator of the one of the meter images.
Description
BACKGROUND OF THE INVENTION
1. Field of the Invention
The present invention relates to a method, an apparatus and a
display used for monitoring the operation of complex process
control systems such as those used in nuclear power plants, and,
more particularly, to a display which summarizes the status of
functions or processes in the complex system in a manner which
allows the operator to easily identify which processed require
attention in the order in which attention should be given and at
the same time monitor desired system parameters at a more detailed
level to anticipate changes in status.
2. Description of the Related Art
Modern process control systems being installed in complex plants
such as nuclear power plants do an excellent job of automatically
or semiautomatically controlling a complex process within preset
parameters even when there are disturbances in the plants. These
control systems are capable of terminating the process when an
emergency occurs; however, it is still desirable, and required in
many situations, that a human operator make control decisions
especially in emergency situations. It is possible for the modern
automatic process control system to shut a plant down when an
emergency occurs; however, human intervention may be able to solve
a particular problem without plant shutdown if the problem is
identified early enough in a degenerating situation.
In many complex processes, the operator is confronted with a vast
amount of information that must be analyzed before appropriate
action can be taken. In attempting to filter out unnecessary
information and provide the operator with information appropriate
to the problem, several different monitoring and display systems
have been created. One such type is a status tree display system
which illustrates the state of the process using a decision tree. A
decision tree is a tool which defines a problem in terms of various
combinations of parameters states which can occur. The parameters
are analyzed sequentially with the active path through the tree
being determined by the states of the parameters being assessed. An
example of such a monitoring and display system is illustrated in
U.S. Pat. No. 4,552,718 by Impink, Jr. et al.
Various methods of summarizing the results of status tree analysis
have been proposed including what is called a status tree bar chart
8 which depicts the state of the functions being monitored, as
illustrated in FIG. 1. This proposed status tree bar chart aligns
one end of the bars on one side of a display and uses both color
and words to identify functions and the degree or nature of the
goal dissatisfaction by a particular process function. That is,
words are used to name the function and describe the state whenever
the function is not in a satisfactory state. The alignment gives
the impression of continuous bars as well as the measurement of
something with respect to a reference point. Since reference
measurement points are not appropriate in such status displays,
this aspect of the proposed display causes considerable confusion.
The order of the bars on the proposed display of FIG. 1 is
arbitrary and the functions which are in a satisfactory state are
merely identified by an existing bar without words indicating which
bar corresponds to which function. The proposed display requires
the operator to memorize the location of each bar.
SUMMARY OF THE INVENTION
It is an object of the present invention to provide a display that
identifies each function at all times, indicates discrete function
states by summary status bare width, color and wording and clearly
indicates a discrete state display.
It is another object of the present invention to order the function
status bars by priority allowing state ties to be resolved by
vertical priority and to associate with each state a procedure for
solving the problem.
It is a further object of the present invention to indicate the
quality of data used to determine the state in association with
each state bar thereby providing a confidence level indicator for
the system state display.
It is an additional object of the present invention to provide, on
the same display as the summary status bars, meter images of the
parameters for operator selected functions associated with the
active path in the decision tree used to produce the associate
status bar to allow the operator to anticipate changes in
status.
It is still another object of the present invention to indicate the
state ranges on the meters as well as data quality.
It is still a further object of the present invention to provide a
second level status tree display accessible through cursor poke
points which displays the status tree for the associated function
and highlights the active path so that the operators can review the
decision making process producing the current function state.
It is still another object of the present invention to provide a
second level status tree display which indicates parameter values
as well as data quality.
It is also an object of the present invention to also provide a
second level display which will allow the operator to continue to
monitor the summary display using a miniature version of the
summary status bar display depicted on the second level
display.
It is a further object of the present invention to provide a poke
point cursor accessible third level display which allows the
operator to look at the individual sensor values which are used in
the computations as well as to continue to monitor the summary
status bar display via the miniature version thereof.
It is an object of the present invention to provide a system of
displays which allows an operator to confirm the validity of
decisions as to process control system state made by the
system.
The above objects can be attained by a display, a display method
and an apparatus which produce a summary display which depicts
function states using discrete state status bars centered in one
window of a two-window display. The discrete states are emphasized
by the centering of the status bars as well as by reference lines
which indicate the state of each function. Priority among functions
is indicated by vertically ranking the bars with the most important
function appearing as the top bar. Each bar includes the name of
the function, the state of the function and a procedure name for a
procedure which can be used to solve the problem associated with
the current state. The discrete function state is also indicated by
the width of the bar and the bar color. On both sides of each bar,
a data quality indicator appears when the quality of the data used
to determine the current state is less than the best.
The other window on the summary display includes meters for
parameters used to calculate the current state along the active
path of the status tree for the most relevant function as
determined by the operator. Each meter indicates not only the
current value of an active path parameter but also the ranges of
the different function states as they correspond to the parameter
being displayed by the meter. The meters also carry the data
quality indicator.
A poke point/cursor accessible second level display is provided
which depicts the status tree with the active path highlighted
allowing the operator to review the decision process made by the
status tree in determining the current state of a selected
function. The second level display also indicates the values of the
parameters used in the decision-making process. The second level
display further includes a miniature version of the summary bar
portion of the highest level display that allows the operator to
continuously monitor the system summary status while reviewing the
second level display.
A third level display, also poke point/cursor accessible, is
provided which depicts detailed information on the sensor values of
the parameters being monitored as well as the exact question being
answered by the process control monitoring system for a desired
portion of a function. This third level display also includes a
miniature version of the summary status bars.
These together with other objects and advantages which will be
subsequently apparent, reside in the details of construction and
operation as more fully hereinafter described and claimed,
reference being made to the accompanying drawings forming a part
hereof, wherein like numerals refer to like parts throughout.
BRIEF DESCRIPTION OF THE DRAWINGS
FIG. 1 depicts a prior art summary display;
FIG. 2 is a schematic diagram of a pressurized water reactor to
which the present invention can be applied;
FIG. 3 illustrates a top level or summary status display in
accordance with the present invention;
FIGS. 4A and 4B illustrate a second level status tree display in
accordance with the present invention;
FIGS. 5 and 6 illustrate the status tree of FIG. 4 including data
quality indicators;
FIG. 7 depicts a third level parameter display in accordance with
the present invention;
FIG. 8 illustrates software modules for producing the displays in
accordance with the present invention;
FIGS. 9A and 9B illustrate one example of a status tree computation
program;
FIG. 10 is a flowchart of the display generator software which
produces the status summary bars;
FIG. 11 is a flowchart of the meter selection software;
FIG. 12 is a flowchart for the procedure which draws the meters of
the summary display;
FIGS. 13A, 13B, and 13C illustrate the flowchart for producing the
second level status tree display; FIG. 14 depicts a flowchart for
production of the third level sensor value display; and
FIG. 15 illustrates an equipment configuration for performing the
procedure of the present invention.
DESCRIPTION OF THE PREFERRED EMBODIMENT
The present invention can be used for various types of complex
nonlinear process control systems such as chemical plants, however,
it will be described with reference to a nuclear generating
unit.
As illustrated in FIG. 2, a nuclear generating unit includes a core
10 inside a reactor vessel 12 which is housed in a containment
structure 14. A coolant heated by a nuclear reaction is circulated
through pipes to steam generators 18 by pumps 20. A pressurizer 16
maintains the pressure in this primary loop. Water introduced into
the steam generators 18 is converted into steam and used to drive a
turbine 22 which drives an alternating current generator 24. The
vitiated steam from turbine 22 is collected in condenser 26 and
recirculated to the steam generators 18 by pump 28. Monitoring and
control of the nuclear generating unit occurs at various points
throughout the plant as indicated by the circles enclosing a letter
representative of the type of reading which is taken at the
particular location, in which F represents flow, L represents
level, P represents pressure, R represents radiation and T
represents temperature. The sensor values produced by the relevant
sensors are monitored by a data acquisition system and used to
produce status displays for the human plant operators.
One of the jobs of the plant operator is to monitor the status of
critical safety functions of a nuclear power plant such as core
cooling. In past implementations, operators have manually evaluated
the status of critical safety functions by following the status
trees in a sequential manner and integrating the results of all the
trees in their minds. To relieve the operators of this time
consuming task, many utilities are considering the automation of
the status tree logic. Thus, a need has arisen for providing the
operator with a summary display which not only depicts the status
of the entire system but also illustrates the value of various
parameters being monitored. Such a display in accordance with the
present invention is illustrated in FIG. 3.
The summary status level display, in accordance with the present
invention, as illustrated in FIG. 3, includes a left-hand window 30
which depicts the summary status of the system being monitored and
a right-hand window 32 which depicts process control parameters in
the active or critical path of a function state computation
selected by the operator. It is also possible to have the meters
associated with the active path automatically appear on the screen
whenever a particular event occurs such as in a trip (automatic
shutdown) of a nuclear power plant. The summary status bar display
side includes two main features: 1) a state reference grid 34 which
visually defines the four levels each function can occupy using
reference lines 36, and 2) status bars 38-48 which depict the state
of each function being monitored. Each status bar, as depicted in
FIG. 3, varies in width in accordance with the function state. For
example, status bar 38 indicates satisfaction of the subcriticality
function, is the smallest sized bar and is green in color. Status
bar 40 indicates a not satisfied core cooling function and is
yellow in color. The next highest goal violation is indicated by
status bar 44 which indicates a severe challenge to integrity and
is orange in color. While the highest goal violation bar is the
jeopardy bar as illustrated by bar 42. Status bar 46 which has its
border removed, as indicated by the dashed line, indicates that the
status of containment cannot be determined because of bad or
inadequate data, which will be discussed in more detail later.
As can be seen from FIG. 3, the status bars not only vary in width
in accordance with the goal but also color and wording. The
centering of the status bars and the vertical reference lines
clearly differentiates the display as a discrete state level
display in contrast to a continuous level bar graph type display.
Each status bar, for example, the core cooling status bar 40
includes three regions or display fields including: 1) a function
name region or field 50 in which the function name is a constant,
2) a written goal status description region or field 52 in which a
description of the status of the function is written, and 3) a
procedure indication region or field 54 which indicates a procedure
name for a procedure which can be used to solve the problem
associated with the dissatisfied goal. Both the written description
region 52 and the procedure indication region 54 are variable
depending upon the status of the particular function.
Whenever the data used by the process control monitoring system is
so bad or in error that it is impossible for the status
determination algorithms to compute a status, a status bar as
illustrated by status bar 46 is displayed in the left window 30 of
the top level display of FIG. 3. The status bar has its outline
color removed and the wording "status indeterminate" is substituted
in the state description region 52. This type bar is displayed, for
example, when all the sensors necessary to determine whether the
containment is operating properly are malfunctioning.
Associated with each status bar is a data quality indication region
56 as particularly illustrated along side heat sink status bar 42.
The poor "P" data quality indication region or indicator 56 for the
heat sink bar indicates that the computation for determining heat
sink function status can be made even though all the sensors
associated with determining the function are not operating
properly. For example, if two out of three sensors for one portion
of the heat sink function are not properly operating and the system
performs calculations based on the highest value among the three
sensors, then the data quality is poor since the value of one
sensor cannot be determined. The data quality computations will be
discussed in more detail hereinafter.
As discussed above the summary status bar display of window 30
takes advantage of the natural stereotype of importance associated
with larger objects are more important than smaller objects and
items near the top of a stack are more important than items near
the bottom resulting in a display with two dimensions of priority.
A display taking advantage of these natural stereotypes is a more
informative display.
The right-side window 32 on the top level display of FIG. 3 is used
to display parameter meters indicating the values of parameters
used to determine function status which the operator wants to
monitor more closely. The meters are selected by activating a
cursor poke point associated with the status bar of interest. By
putting the selectable meters on the right-hand side of the
display, the visual image again takes advantage of a natural
stereotype where people read from left to right, that is the most
important display, the summary status bar display, is read first
when confronted with the combined display. Each meter corresponds
to one or more decision points in the status tree associated with
the active path or critical path of the function being monitored,
that is, all the meters associated with the status tree are not
shown, only the active path parameter meters. The number of meters
displayed can range from one to six for the Westinghouse owner's
Group Critical Safety Function Status Trees depending on the
corresponding status tree and, of course, the number would vary if
the present invention is applied to a different process.
Each meter includes a meter background or template 58 which
includes the outline of the meter, the meter ranges and the range
description. For example, meter 60 has a range of from 0 to 100 for
percent level of the highest steam generator average level. Inside
each meter template is an actual value indicator 62 which indicates
the current value of the parameter monitored within the range
indicated on the right side of the meter. A diamond shaped
indicator 62 is preferred because it indicates that the value
corresponds to both sides of the meter and when the indicator
reaches the bottom of the meter the indicator does not disappear as
would occur if a line or bar were used as the indicator 62. To the
left of the meter are colored brackets 64-72 which indicate the
range of the various status levels associated with the parameter
being monitored. In meter 58 four ranges are indicated. The top
range 64 indicates the jeopardy level and is yellow in color the
next lower range 66 can indicate the not satisfied range and would
be yellow in color or the satisfied range and be green in color.
Between the two highest ranges 64 and 66 is a threshold indicator
72 that depicts the point at which the status of the function will
change in dependence upon the particular parameter being monitored.
The third level bracket 68 and the bottom level 70 can be the
jeopardy (red), not satisfied (yellow) or the satisfied range
(green). The color chosen depends on the remaining values at the
tree nodes. The bottom range 68 can belong to range 66 or 70
depending on which set point is currently applicable, 6% or 34%.
However, a meter could include brackets for all the colors in the
system that indicate determined states. Between all the ranges are
the threshold change indicators. Each of the colored status range
brackets associated with the meters can change in range or size
because status is determined in accordance with many parameters and
as a result the threshold points between ranges will also move up
or down depending on how the particular parameter being monitored
is affected and affects other parameters.
Below each meter in FIG. 3, for example, meter 58, is a constant
description region 74 which describes the parameter being monitored
by the associated meter 58. Below the description is a digital
read-out region 76 indicating not only the actual parameter value
but also the particular equipment being monitored, when more than
one piece of equipment can be used to obtain the monitored
parameter, for example, steam generator loop X.
Meter 78 has the same features as meter 58, however, the actual
level indicator 80 in meter 78 indicates that the data quality of
the measured parameter is poor. In addition to the poor and
indeterminate or bad data quality indicators a manual indicator 82
as illustrated in meter 84 can be used. This indicator is produced
both on the meters and in the summary display of window 30 whenever
the operators have entered a manual value for the sensor values
used to compute the parameter value. The manual indicator 82
indicates to the operator that the sensor values for this parameter
must be updated manually before the reliability of this meter can
be accurately determined. For example, when the sensor values for
water level in pressurizer 16 of FIG. 2 cannot be determined by the
sensors because of some malfunction, the operator may be able to
telephone the plant and receive a visually determined level value
from a plant technician. The operator then can enter this manual
value into the status tree calculations and obtain a status
calculation for inventory, even when the sensors associated with
measuring inventory are malfunctioning. Meter 86 depicts a meter in
which the parameter being monitored cannot be determined because
of, for example, bad data. In this meter the word indeterminate is
written in place of the indicator. In addition, it is also possible
to change the bracket to a color such as magenta to also indicate
bad data. Meter 86 also illustrates a meter for which the entire
parameter range is not relevant for the application and, thus, only
the relevant portion is presented.
The combination of the summary status window 30 with the meter
window 32 provides a more powerful display to the operator which
allows the operator not only to monitor the overall status of the
system but to also monitor in detail the parameters that affect the
active decision tree path or current state of a selected function.
Such a display provides the operator with much more information
than either the summary display alone or the meter display alone as
well as much more information than the display of a status tree
alone. For example, in an active version of FIG. 3, the operator
can monitor the highest steam generator average level (meter 60)
for the heat sink function (status bar 42) and determine that some
progress is being made in correcting a heat sink problem by
visually watching the movement of the actual level indicator 62.
Thus, as the indicator moves up or down within the meter 60, the
operator can determine rapidly whether his efforts are solving the
problem as depicted by the position of the pointers 62 with respect
to the range brackets (64-70) as well as continue to monitor the
status of other critical safety functions within the plant using
the summary status bar display window 30.
If the operator needs to review the status tree associated with,
for example, the heat sink function, the operator can activate a
cursor poke point on the display of FIG. 3 and an actual active
status tree diagram such as illustrated in FIG. 4 for the heat sink
function will be displayed. The status tree includes all of the
decision tree (binary decision points) and branches as well as the
questions used to determine the status of the particular function.
For example, the status tree for heat sink includes six decision
nodes 200-210. Each decision node, for example decision node 200,
includes a question region or field 212 which depicts the fixed
question being asked about the parameter being monitored. Also
included in the node are answer fields 214 and 216 of which the
particular binary answer at the decision node is highlighted by
showing the answer in reverse video. Also included is fixed wording
description region or field 218 along with a variable parameter
value region 220. The provision of the value region 220 allows the
operator to verify the answer to the question and attain incite as
to the reason for the highlighted answer.
The connections between the decision nodes are illustrated by
branch lines 222-248. As can be seen by the status tree depicted in
FIG. 4, the active path or critical path of the decision tree is
highlighted by thickening the line. It is possible, of course, to
provide a color highlighted line for the active path. The remaining
branches are merely indicated by non-highlighted branches. Located
at the end of the last branch 228 in the active path is a
termination node highlighted box 250 surrounding a status
description and a procedure name which should be followed to solve
the problem associated with the active state. The remaining
termination nodes 252-260 are not highlighted. The decision tree
can be used by the operator to investigate which parameter, if
changed, will have the greatest effects on the current state of the
function. However, typically the meter display would be used for
the investigation function and the decision tree to confirm state
logic.
In one corner of the second level status tree display is a
miniature version 262 of the summary status bar display of FIG. 3.
This display is substantially identical to the larger summary
status bar display in that the relative sizes of the status bars
are maintained along with the data quality indications. However,
the wording in each bar indicating an unsatisfied goal only sets
forth the procedure that should be followed to solve the particular
problem, that is, the function name and state wording are absent.
The provision of the miniature display 262 all on the second level
status tree displays allows the operator to continue to monitor the
overall system status while investigating a particular problem. For
example, when the operator examines the second level display of
FIG. 4, the operator can determine that the jeopardy state can be
changed to the not satisfied state by increasing the total
auxiliary feed water flow to all the steam generators to greater
than 377 gallons per minute or by forcing the narrow range level of
at least one steam generator to be greater than a certain
percentage value while also monitoring system status.
When the data quality of a particular parameter is not the best,
the decision tree will depict poor data quality as illustrated in
FIG. 5. By following the "P" code branches, in this example node
264, the particular parameter values in error or out of range can
be located. The poor parameter values are indicated by a data
quality indicator "P" associated with the value in the parameter
value region 220. This will allow the operator to visually
determine which sensors should be addressed by maintenance
personnel to improve data quality. If manual data had been
substituted for an actual sensor value, the "P" lines of FIG. 5
would be replaced by "M" lines. FIG. 6 illustrates an example of a
display when bad data is encountered by the status tree computation
routines.
If the operator needs to examine the sensor values which go into a
particular decision tree node or parameter computation, the
operator can poke point access a third-level display as illustrated
in FIG. 7. This display includes fixed question fields 290
displaying the question of the node being investigated along with
answer fields 292 corresponding to each question. In addition, also
included are representations of the actual calculation used to
determine the actual value. For example, in FIG. 7, the steam
generator A pressure equation representation 294 indicates that
steam generator A pressure is a three-channel average. Within this
representation region 294 is the actual computed value 296.
Underneath the equation representation are actual sensor value
representation fields 296 for the actual values of sensor pressure
for steam generator A pressure. To the right of the sensor value
fields are fixed wording tag fields 298 which provide an
identification number for the particular sensor being monitored. If
the sensor is malfunctioning, X's are substituted for the
appropriate actual values on this display. In general, other
calculations for other parameters and decision nodes in a system
such as a nuclear power plant monitoring system are as simple as
the equations represented in FIG. 7.
As can be seen on the lower left-hand side of FIG. 7, the miniature
version 262 of the status bar display of FIG. 3 reappears on this
third-level display. The reproduction of the miniature version of
the status bar display allows the operator to continue to monitor
overall system status as he investigates a particular problem,
thereby providing the benefits of a summary system display
throughout problem investigation.
The combination of displays discussed above allows the operator to
continuously monitor system status using a display having the
priority of the various critical functions indicated by the order
of the display status bars and the discrete states of each function
status indicated not only by color and wording but also by
reference lines. While the particular procedure that can be used to
solve the problem is carried with the displays as the operator digs
more deeply into the system to solve a particular problem. That is,
the present invention provides a display with substantially more
information available to the operator while maintaining the
high-level summary type display providing substantial benefits over
prior art displays when solving critical problems.
FIG. 8 depicts the conceptual division of the software of the
present invention into various modules. Of course, depending on the
size of the computer and cycle time for the displays, these modules
could be combined into larger modules or subdivided into even
smaller modules. The first module 400 represents the data
acquisition software which is used to read sensor values and store
these values in a centralized dynamic data base. The data
acquisition software can be provided by one of ordinary skill in
the art or purchased from a vendor such as Digital Equipment, Inc.
The programming language for the data acquisition module is
preferably an assembly language since a real time data sampling
task is involved.
The design of the initial data quality module 402 is also within
the ordinary skill in the art. This module does simple computations
to determine the initial quality of the data gathered by the data
acquisition module. For example, if a particular sensor among
triply redundant sensors is malfunctioning, the initial data
quality module will tag the data block containing the inoperative
sensor value with an indicator indicating that the data is of
questionable quality. Another example is if a particular sensor
value is out of an acceptable range, the sensor value field will be
tagged with a bad quality indicator. The quality determination
module 402 will compare the sensor value with the range and if out
of range add a bad data tag. The particular data quality
computations or algorithms are dependent on the type of system
being monitored and the redundancy of the sensors in the monitoring
system. A preferred language for such computations would be an
assembly language which would allow quick data quality checks.
The third module 404 is the status tree computation module. This
module can be programmed in a language such as Fortran from a
flowchart as illustrated in FIG. 8 to be discussed in more detail
later. One of ordinary skill in the art could provide these
hardcoded procedures as discussed in detail in U.S. Pat. No.
4,552,718 previously mentioned and incorporated by reference
herein. However, as more sophisticated rule-based or artificial
intelligence type analysis systems are integrated into process
control systems, the application of such inference software as the
status tree computation module would be appropriate. This approach
would increase the flexibility of the system in adapting to
additional new or critical functions.
The display production module 406 is a procedure which actually
produces the display and uses, for example, overlay techniques. The
display creation procedures will be discussed in more detail later;
however, well known procedures such as the performing a byte
transfer overlay or a cursor poke point acknowledgement will not be
discussed in detail since one of ordinary skill in the art can
provide the details of such procedures.
FIG. 9 depicts a typical flowchart for a critical function status
computation and is the flowchart for the heat sink computations
illustrated in FIG. 4. Symbol 412 represents the calculation of a
logical parameter ADVCTMT which is the determining factor for using
either 6% or 34% as the setpoint for decision nodes 200 and 210 in
FIG. 4. The details of this calculation are with the ordinary skill
the art. A comparison of this flowchart with FIG. 4 indicates that
the decision steps in the software parallel the decision steps of
the status tree thereby making the conversion of status trees into
appropriate status tree analysis software very simple. At each
decision level, for example, decision block 418, a branch is taken
which corresponds to a decision node in the status tree, in this
example corresponding to decision node 200 in FIG. 4. Block 420
represents a time delay which allows the auxiliary feedwater pumps
to start up completely before decision node 204 in FIG. 4 is
evaluated. Without this time delay, an operator would be
immediately directed to procedure FRP-H.1 as indicated by the red
status of endpoint 250 in FIG. 4. The multiple path branch of FIG.
4 which allows both answers to node 200 to follow a path through to
node 204 is illustrated by the branch pointer "2" in FIG. 8 which
joins blocks 422 and 426. As can be seen by the output blocks, for
example 428, the output equals the output color as well as the
procedure name. The color represents the status when mentioned in
later discussed flowcharts.
To produce the displays of the present invention, one method
involves loading a display background into a display generator
memory using a well-known byte transfer technique and overlaying
the background each time the display changes also using a byte
transfer technique. For example, if the memory address of the
beginning of a fixed text field in the generator memory is known
the text could be transferred to the memory while writing over the
contents of each memory location following the known address. Such
overlay techniques are well known in the display creation art.
There are other methods of creating such displays such as
line-by-line creation methods, however, the overlay method, even
though it consumes larger amounts of memory, is generally faster
and is preferred for real time systems, such as in a nuclear power
plant. In the discussion below, whenever the loading of a
background display or the overlay of a particular portion of a
display is indicated, one of the well-known overlay procedures is
used to load the appropriate portions of a display generator
memory.
FIG. 10 illustrates the flowchart of the procedure for creating the
summary status bar display of FIG. 3. The software represented by
FIG. 10 can be executed continuously or cyclically as determined by
the refresh or scan cycle for the data acquisition module. The load
display background step 462 loads the fixed wording regions 50 of
the status bar as well as the reference lines 36 for the display
into the display generator. The critical safety function status,
the procedure numbers and data quality for all the functions are
retrieved in step 464 from a memory associated with the previously
executed modules or from, for example, a common or global data
base. If all the bars have not been processed 466, a determination
is made 468 as to whether or not the data quality is bad. If the
data quality is bad, the word status indeterminate is overlaid 470
in place of the bar. It is also possible to create a border for a
bad data quality bar having a color such as magenta which would
indicate bad data by including an appropriate overlay at this
step.
If the data quality is not bad, the appropriate size bar, color and
procedure name are overlaid 472 in the appropriate location in
display generator memory. If the data quality associated with this
particular critical safety function is poor the appropriate
indicator "P" is overlaid 476 in the appropriate data quality
indicator region 56 for the respective bar. Similar steps are
performed 478 and 480 for the manual data quality indicator. When
all bars have been processed, the system can move into a wait state
in which the procedure determines whether a status display is
appearing 482 at any terminal. If the status or data quality has
changed 484 for an active status bar display, the display
background is again loaded to clear the display, of for example an
oversized bar, and the process of creating the bars is again
executed.
The process for creating the miniature display associated with the
second and third-level displays is substantially identical to the
process depicted in FIG. 10 and will not be further discussed for
simplicity purposes.
Depending upon the particular cursor poke point activated by the
operator, the meters for the right-hand window of FIG. 3 are
selected in accordance with the flowchart of FIG. 11. This module
only selects those meters associated with the corresponding status
tree that are along the active path, and calls another routine
(FIG. 12) that actually draws the meters, that is, as the active
path in the status tree changes the meters appear and disappear in
correspondence thereto. If the top level display is currently
appearing 492, the tree logic for the desired display and terminal
nodes 494 are retrieved.
The tree logic can be any of a number of data structures used
representing the tree with the preferred structure being a linked
list, that is, each record in the data structured preferably
represents a node and includes the actual fixed text fields
question asked or the status if a terminal node. Also included in
each record is a logical representation of the question, a logical
representation of the answer, the fixed test of the parameter
description or procedure name, the parameter values and pointers to
each subsequent record in the tree along with a node type indicator
that indicates whether the record represents a terminal node or an
intermediate node. Various other flags and indicators such as a set
point change flag can be included in each node record as discussed
hereinafter. To follow the tree logic to obtain needed field
contents, the software only need to follow the pointers using the
answer representation to determine which branch to follow. The tree
traversal procedures are well known to those of ordinary skill in
the display art.
Next, the appropriate meter for the parameter in the first question
in the tree is drawn 496 by the procedure depicted in FIG. 12.
Then, a decision is made 498 as to whether there are subsequent
questions in the tree by examining the node type indicator for the
current node, if there are no subsequent questions the status of
the particular critical safety function is written 500 at the top
of the display of FIG. 3. If there are additional questions, a
determination is made 502 as to whether the meter for the parameter
of the next question is already on the display. If not, the
appropriate meter is drawn 504.
FIG. 12 is a flowchart of the meter drawing procedure. This
procedure not only draws the meter but also determines the
appropriate color and the problem solving procedure for the
brackets of the meter by following the branches of the tree until a
terminal node having an associated color is encountered. This
routine also inserts the actual value indicator 62 and accompanying
data quality indication symbols.
First, a determination 512 is made as to whether the set points for
this parameter change with sensor environment conditions. This
determination can be made by examining the parameter change point
flag in the current node record. If the answer to this is yes,
another determination is made 514 as to whether or not the sensor
environment conditions are adverse 514 using a similar flag in the
current node record. If the sensor environment conditions are not
adverse normal set points are retrieved 518 from the node record.
Otherwise, adverse function set points are retrieved 516 from the
node record. When the set points are determined, the appropriate
brackets with appropriate colors can be overlaid 520 onto the meter
background.
At this point, the appropriate indicator for data quality as well
as current parameter value is overlaid or written into the meter
image along with the digital values in the description section 76
below the meter 520. If all the brackets are assigned a color 522,
the process terminates 534 and if all the brackets are not assigned
a color a determination is made 524 as to whether the current value
of the parameter is in the range of this bracket. If the answer is
yes, the color associated with the current status is overlaid. If
the parameter is not in the range of the particular bracket, a
determination is made 528 as to whether additional questions nodes
must be traversed to reach a terminal node to indicate the color of
the bracket. If no additional questions are necessary, the color of
the terminal node (status) is overlaid on the bracket 530. If
additional questions must be answered, the nodes or records of the
tree structure are traversed following the path, that would be
active if the parameter crossed into the range of that bracket,
using pointers until a terminus is reached at which the color of
the bracket can be determined and overlaid.
If the operator accesses the second level display for a particular
critical function, a procedure in accordance with the flowchart of
FIG. 13 is executed. This procedure simply loads the background
into the display generator and then overlays the display with the
active data at the proper locations.
First, the background for the display is loaded 542 which includes
the decision node diagrams (the boxes and branches and terminal
boxes) and the fixed text including procedure names and questions.
Next, the active data for the parameter values are retrieved 544
and loaded into the appropriate variable value fields. The answer
to the first question is retrieved 546 and decisions 548-558 are
made as to the quality of the data. If the data is poor, a poor
data flag 562 is turned on and the branch associated with the
question answer is overlaid 564 with the poor question code as
illustrated in FIG. 5. If the data is manual data, the manual data
flag is turned on 566 and the appropriate manual data code is
overlaid 568. If data quality for all data is good, the answer is
highlighted and the subsequent branch is overlaid 560 with a thick
or highlighted line.
If the data quality is not bad (that is, poor, manual, or good) a
determination is made 570 as to whether there are subsequent
questions in the tree by examining the node type indicator. If
there are subsequent questions, the answer to this question is
retrieved 572 along with its data quality and the loop for
determining the appropriate data quality code, if any, is again
executed. If there are no subsequent questions in the tree along
the active path, the active path terminal box is turned 574 to the
appropriate color. A determination is made 576 as to whether the
second level display is still active and if so the current data is
erased and the process is started again.
If the answer does have bad data quality, the bad data flag is
turned on 580 and one is added to a bad data quality count. The bad
data quality line code is overlaid 584 on the subsequent branch to
this question. Next, an analysis is made 586 as to whether there
are subsequent questions in the yes branch for the previous node.
If the answer is no then a determination is made 588 as to whether
there is a subsequent question in the no branch for the previous
question. If the answer to the questions in boxes 586 and 588 is
yes, the answer for the current question and data quality are
retrieved 572.
If the bad data quality flag is on, the question answers are
highlighted and the subsequent branch is overlaid 590 with the bad
data quality code. Once again, a determination is made 592 as to
whether there is a subsequent question in the tree. If there are no
subsequent questions, then the answer for the previous branch and
its data quality are retrieved 594 and a determination is made 596
as to whether the prior question has a bad data quality. If the
answer is yes, then a determination is made as to whether this
branch is associated with a yes answer. If not, one is subtracted
600 from the bad data quality count and a determination is made 602
as to whether the bad data quality count is equal to zero
indicating that all branches of the tree have been processed. If
the answer is yes, the system determines 576 whether the second
level display is still active and recycles through the process.
If the operator accesses the third level display, a procedure in
accordance with FIG. 14 is executed. First, the background
information for this display which includes the question, and the
equation representations is loaded 612 into the display generator.
Next, the active data for this function is retrieved and placed 614
in appropriate locations of the display. Then a determination is
made 616 as to whether the third-level display is still active. If
the third-level display is still active, the active data is erased
618 from the display and the process starts again.
Many hardware configuration are appropriate for a system which
processes and produces displays in accordance with the present
invention. One preferred for a nuclear power plant is illustrated
in FIG. 15. Sensors 700-702 are accessed by a data acquisition
computer 704 such as a Gould concept 32 6750 series machine and the
collected data is transferred to a data quality and status tree
computation computer 706 which can also be a Gould concept 32 6750.
This machine 706 performs the computations of FIGS. 8-14 and sends
appropriate overlays to the display memory of a display generator
708 such as a Matrox display generator from Matrox Electronics
Display Systems Model SX900. Typically, the display generator would
store the background with the overlays being created by computer
706. A keyboard 17 is used by the operator to access both the
computation computer 706 and display generator 708 through a
keyboard processor such as an Intel 28610. The display generator
produces an image on a CRT display such as the 8835 produced by
Aydin. Other configurations are appropriate for executing the
method of the present invention and consideration must be given to
the cycle time required as well as the volume of data which must be
gathered and analyzed.
The system of displays as previously described allows the operator
to move through the displays to confirm the validity of the inputs
to the system, the logic used to determine system status and the
output status determined by the process control monitoring system
and to thereby increase his confidence level in the validity of the
output. This allows the process control monitoring system when
operating at a high confidence level to positively contribute to
correcting plant abnormalities. The system also allows the operator
to monitor over all status while at the same time watch and
anticipate status changes.
The many features and advantages of the present invention are
apparent from the detailed specification and, thus, it is intended
by the appended claims to cover all such features and advantages of
the invention which fall within the true spirit and scope thereof.
Further, since numerous modifications and changes will readily
occur to those skilled in the art, it is not desired to limit the
invention to the exact structure, operation and method illustrated
and described, and accordingly all suitable modifications and
equivalents may be resorted to, falling within the scope of the
invention.
* * * * *